Process Mining and Learners' Behavior Analytics in a Collaborative and Web-Based Multi-Tabletop Environment

Interactive tabletops alone cannot automatically analyze and interpret students' digital footprints (event logs). Moreover, the final artifacts created by groups provide imperfect information about each individual's contribution to the group task. This research is divided into two main parts. In the first part of the study, a quantitative survey was conducted in order to identify the most significant indicators affecting the collaboration process in an online and networked context-aware multi-tabletop environment. In the second part of the study, several process mining techniques such as social network mining, basic performance analysis, role hierarchy mining, and dotted chart analysis were used with the purpose of increasing the instructor's awareness/knowledge about the collaborative dynamics in each group. The empirical findings showed that the levels of symmetry of actions (or similarity of tasks) and symmetry of roles (or low division of labor) were much higher in the high-performance groups compared with the low performance groups. Consequently, high performance groups showed increased tendencies to work on the same range of actions ‘together'. Quite the opposite, low performance groups showed increased tendencies to work on a dissimilar range of actions ‘individually'.